Optimization Methods for Scaling Services under High Load Conditions

This article explores methods for optimizing the scaling of cloud services under high-load conditions. The study draws on both traditional approaches to load distribution—vertical and horizontal scaling, load balancing algorithms, caching techniques—and modern technologies, including containerization, serverless computing, edge computing, and hybrid models that combine heuristic algorithms with deep learning methods (CNNLSTM). A comprehensive literature review helps identify key performance metrics: response time, total makespan, number of virtual machine migrations, algorithm runtime, and energy consumption. The findings demonstrate that the integration of hybrid models contributes to reduced response time and overall task completion time, fewer migrations, and improved energy efficiency compared to traditional approaches. The practical applicability of the described methods is confirmed by their potential implementation in enterprise cloud infrastructures that must meet requirements for fault tolerance and system adaptability. This is especially relevant for scalable services operating under dynamic loads in modern business and IoT environments. The article will be of interest to researchers and practitioners in the field of distributed computing and high-load systems, including IT infrastructure architects, DevOps professionals, and engineers building resilient and scalable services.